Image crashes on startup with ORA-03113: end-of-file on communication channel about oci-oracle-xe

[gvenzl]

Thanks a lot, @HenrikDK!

Oh yeah, of course :)

It appears that the database terminates deliberately because one of its background processes died:

2021-11-28T20:23:28.437947+00:00
PMON (ospid: 35): terminating the instance due to ORA error 12752
Cause - 'Instance is being terminated due to fatal process death (pid: 8, ospid: 49, SA00)'
2021-11-28T20:23:28.448006+00:00
System state dump requested by (instance=1, osid=35 (PMON)), summary=[abnormal instance termination].
System State dumped to trace file /opt/oracle/diag/rdbms/xe/XE/trace/XE_diag_54_20211128202328.trc

Looking at the docker info output I'm taking a wild guess here that the container does not have enough memory to run:

 Kernel Version: 5.10.47-linuxkit
 Operating System: Docker Desktop
 OSType: linux
 Architecture: x86_64
 CPUs: 2
 Total Memory: 1.94GiB
 Name: docker-desktop

The database alone is trying to allocate 2GB for its SGA and PGA memory areas, that would not leave enough memory for any of the processes running and my guess is that the kernel Out of Memory Killer starts to kill processes.
Could you try to increase the memory for Docker to say 3GB (perhaps 2.5GB should already be enough) and see whether this issue still occurs?
Also, if you still have the container on your machine, could you also please get the trace file /opt/oracle/diag/rdbms/xe/XE/trace/XE_diag_54_20211128202328.trc out of it?

[gvenzl]

Hey @HenrikDK,

Unfortunately, there is not enough information in the docker logs output to determine the cause of that issue.

Could you please provide me with a docker info and also the last couple of lines in the database alert log inside the container via docker exec oxsql tail -n 50 /opt/oracle/diag/rdbms/xe/XE/trace/alert_XE.log?
That should have more information as to what is going on under the hood.

Thanks,

[HenrikDK]

Hi, sure heres the output,

docker-info.txt

the docker command you provided didn't work since the container was no longer running, but i hadnt deleted the instance so so I did a copy instead:

docker cp f065ff3bc8a1:/opt/oracle/diag/rdbms/xe/XE/trace/alert_XE.log alert_XE.txt

alert_XE.txt

[HenrikDK]

Hmm, thought it might be something like that, I've adjusted the memory to 6gb and will try again.

Here's the trace file you requested:
xe_trc.txt

[HenrikDK]

Looks like its working now, image hasn't exited and I can connect to the DB. Thanks for the help :)

[gvenzl]

Wohoo, there we go, glad to hear!
3 GB should be plenty for one container btw, but if you have the memory, sure 6GB works as well. :)

I should probably add some memory check into the container startup script, will open an ER for this and close this one.

Thanks and thanks for using these images! :)

[cngpc43]

I met the same problem. Here is my Container's log:

2023-09-12 23:32:43 CONTAINER: starting up...
2023-09-12 23:32:43 CONTAINER: database already initialized.
2023-09-12 23:32:43 CONTAINER: starting up Oracle Database...
2023-09-12 23:32:43 
2023-09-12 23:32:43 LSNRCTL for Linux: Version 21.0.0.0.0 - Production on 12-SEP-2023 16:32:43
2023-09-12 23:32:43 
2023-09-12 23:32:43 Copyright (c) 1991, 2021, Oracle.  All rights reserved.
2023-09-12 23:32:43 
2023-09-12 23:32:45 Starting /opt/oracle/product/21c/dbhomeXE/bin/tnslsnr: please wait...
2023-09-12 23:32:45 
2023-09-12 23:32:45 TNSLSNR for Linux: Version 21.0.0.0.0 - Production
2023-09-12 23:32:45 System parameter file is /opt/oracle/homes/OraDBHome21cXE/network/admin/listener.ora
2023-09-12 23:32:45 Log messages written to /opt/oracle/diag/tnslsnr/018ad28d1b65/listener/alert/log.xml
2023-09-12 23:32:45 Listening on: (DESCRIPTION=(ADDRESS=(PROTOCOL=ipc)(KEY=EXTPROC_FOR_XE)))
2023-09-12 23:32:45 Listening on: (DESCRIPTION=(ADDRESS=(PROTOCOL=tcp)(HOST=0.0.0.0)(PORT=1521)))
2023-09-12 23:32:45 
2023-09-12 23:32:45 Connecting to (DESCRIPTION=(ADDRESS=(PROTOCOL=IPC)(KEY=EXTPROC_FOR_XE)))
2023-09-12 23:32:45 STATUS of the LISTENER
2023-09-12 23:32:45 ------------------------
2023-09-12 23:32:45 Alias                     LISTENER
2023-09-12 23:32:45 Version                   TNSLSNR for Linux: Version 21.0.0.0.0 - Production
2023-09-12 23:32:45 Start Date                12-SEP-2023 16:32:45
2023-09-12 23:32:45 Uptime                    0 days 0 hr. 0 min. 0 sec
2023-09-12 23:32:45 Trace Level               off
2023-09-12 23:32:45 Security                  ON: Local OS Authentication
2023-09-12 23:32:45 SNMP                      OFF
2023-09-12 23:32:45 Default Service           XE
2023-09-12 23:32:45 Listener Parameter File   /opt/oracle/homes/OraDBHome21cXE/network/admin/listener.ora
2023-09-12 23:32:45 Listener Log File         /opt/oracle/diag/tnslsnr/018ad28d1b65/listener/alert/log.xml
2023-09-12 23:32:45 Listening Endpoints Summary...
2023-09-12 23:32:45   (DESCRIPTION=(ADDRESS=(PROTOCOL=ipc)(KEY=EXTPROC_FOR_XE)))
2023-09-12 23:32:45   (DESCRIPTION=(ADDRESS=(PROTOCOL=tcp)(HOST=0.0.0.0)(PORT=1521)))
2023-09-12 23:32:45 The listener supports no services
2023-09-12 23:32:45 The command completed successfully
2023-09-12 23:32:46 ORA-03113: end-of-file on communication channel

The trace file is attached
alert_XE.txt

P/s: My docker desktop resources' setting for memory was 6gb already